AI tool predicts people’s risks of disease
An artificial intelligence tool is capable of predicting an individual’s risk of more than 1,000 diseases and forecasting health changes up to a decade early.
Researchers hope future versions of the AI tool could warn clinicians and patients of conditions long before symptoms appear, allowing for the early roll-out of preventive measures.
The Delphi-2M system draws on a modified large language model, the technology that underpins “generative AI” services such as ChatGPT.
Described in the journal Nature, Delphi-2M examines anonymised patient records alongside lifestyle factors such as obesity, smoking and alcohol consumption, as well as age and sex.
From this, it can predict the likelihood and timing of a person developing conditions such as cancer and diabetes.
The system could be used to help health services plan for the future in as little as two years, said Ewan Birney, the interim executive director at the European Molecular Biology Laboratory and a co-author of the study. Deploying the tool clinically might be possible within a decade, he added.
“Our model is a proof of concept, showing that it’s possible for AI to learn many of our long-term health patterns and use this information to generate meaningful predictions,” Birney said.
No comments:
Post a Comment